Do not blindly follow my work, especially in production. There are no warranties I can provide and these are not Proxmox supplied instructions.
A host was renamed recently. The steps on the wiki indicate to rename some files /etc/hosts /etc/hostsname.
https://pve.proxmox.com/wiki/Renaming_a_PVE_node
Additional steps which could be added;
/etc/hosts should contain the new hostname and fqdn where appropriate
Use `hostname new_node` to set the new name
/etc/pve/nodes/old_node/lxc and /etc/pve/nodes/old_node/qemu-server should be backed up and moved to /etc/pve/nodes/new_node/lxc and /etc/pve/nodes/new_node/qemu-server respectively. This will "move" the nodes to the correct location, otherwise the interface looks funny.
Upon fixing their locations the Proxmox api2 wasn't happy. In order to fix this I issued the command `systemctl restart pvedaemon` so that the service running the api2 (because pveproxy doesn't manage the api) would stop responding with "Proxy not allowed (500)". These messages don't appear to be logged anywhere specifically. You can see the 500 errors in the pveproxy logs but no reason given.
I didn't follow the rrdtool file copy which resulted in the loss of some of my statistics. This was not a big deal for me however if you need that data refer to the official wiki. My host is happy now following these steps.
It is my opinion that the node rename wiki could indicate a reboot is required or incorporate my procedure.
Do not blindly follow my work, especially in production. There are no warranties I can provide and these are not Proxmox supplied instructions.
A host was renamed recently. The steps on the wiki indicate to rename some files /etc/hosts /etc/hostsname.
https://pve.proxmox.com/wiki/Renaming_a_PVE_node
Additional steps which could be added;
/etc/hosts should contain the new hostname and fqdn where appropriate
Use `hostname new_node` to set the new name
/etc/pve/nodes/old_node/lxc and /etc/pve/nodes/old_node/qemu-server should be backed up and moved to /etc/pve/nodes/new_node/lxc and /etc/pve/nodes/new_node/qemu-server respectively. This will "move" the nodes to the correct location, otherwise the interface looks funny.
Upon fixing their locations the Proxmox api2 wasn't happy. In order to fix this I issued the command `systemctl restart pvedaemon` so that the service running the api2 (because pveproxy doesn't manage the api) would stop responding with "Proxy not allowed (500)". These messages don't appear to be logged anywhere specifically. You can see the 500 errors in the pveproxy logs but no reason given.
I didn't follow the rrdtool file copy which resulted in the loss of some of my statistics. This was not a big deal for me however if you need that data refer to the official wiki. My host is happy now following these steps.
It is my opinion that the node rename wiki could indicate a reboot is required or incorporate my procedure.
Do not blindly follow my work, especially in production. There are no warranties I can provide and these are not Proxmox supplied instructions.